(with Stefan Grimm, Dominik Grothe, David Schindler, and Simeon Schudy)

 

Companies increasingly make use of team-based organizational structures in complex non-routine tasks. To foster performance in these settings, scholars and practitioners alike have emphasized the potential of leadership. However, its value is difficult to identify in agile and cross-functional teams since leadership is often determined endogenously. To measure the causal impact of leadership, this study randomly encourages teams in a large-scale field experiment to select a leader who consistently motivates or consistently coordinates their team. Both encouragements substantially improve team performance and systematically alter team organization, without reducing the originality of solutions. We provide further data to describe patterns of team interaction in response to variations in leadership.